- The objective is to count the total amount of money in an image.
- To achieve this, binarization is applied to the image.
- Next, some morphological operations are applied to improve the quality of the results.
- Then, the extraction of connected components is performed.
- Based on the shape, exclude objects that are not coins.
- Afterward, a classifier is applied to determine the value of each coin based on its size.
- Finally, the total amount of money is calculated.
 |
 |
 |
| 1. Input image | 2. Binary image | 3. Better image |
 |
 |
 |
| 4. Image with countours | 5. Selected countours | 7. Final result |
- The objective is to have a program that checks which object is in a video which can be a "person", "car" or "other".
- The algorithm calculates the difference between the previous frame and current frame, identifying what changed from the previous to the current.
- Then, it is extracted the active regions.
- Following that, it is classified the region using a median classifier based on aspect ratio.
- Finally, the path followed by the object is drawn. Note: The video is too large to be stored.
 |
 |
 |
| 2. Difference between frames | 3. Obtain countours | 4. Box around the objects |
 |
| 5. Draw path of objects |